David Bancroft’s column, “Ignoring sprawl will not make it go away,” (May 2007) states that recently developed maps of urban cover derived from Landsat satellite imagery of the Chesapeake watershed—Mid-Atlantic Regional Earth Science Center maps or RESAC maps—are unsuitable for monitoring urban sprawl.
We respectfully disagree and, in fact, have used these image-based map products for just this purpose, documenting their utility and providing guidelines on their proper use in peer-reviewed journal publications (See Goetz, S.J., & Jantz, P. (2006). “Satellite maps show Chesapeake Bay urban development.” Eos Transactions AGU, 87, 149-152 and Jantz, P., Goetz, S.J., & Jantz, C.A. (2005). “Urbanization and the loss of resource lands within the Chesapeake Bay watershed.” Environmental Management, 36, 808-825.)
The figure (at right) of change in impervious cover—the built environment—gives a sense of the spatial detail in the RESAC maps. Using these maps across the entire 168,000 km2 watershed, we have estimated 2,485 km2 (895 square miles) of new development between circa 1990 and 2000.
Partly on the basis of these types of urban change assessments, and at the recommendation of the Land, Growth and Stewardship Subcommittee (See Chesapeake Bay Program Office, Land, Growth, and Stewardship Subcommittee, memorandum to the Implementation Committee, Measuring the Rate of Land Conversion in the Chesapeake Bay Watershed, 15 March 2004), the Chesapeake Bay Program has been using the RESAC maps for the last two years as part of their resource lands assessment; see www.chesapeakebay.net/land.htm.
While it is true that the Bay Program Partners have been struggling with how to accurately and effectively monitor harmful sprawl, this is primarily an issue because it is not possible to routinely assess urban change on an annual basis because of the expense.
Nonetheless, it would be a mistake to dismiss the utility of the RESAC or other satellite-derived maps for monitoring urban change—or, by extension, its impacts on resource lands—simply because the maps can only be produced at approximately five-year intervals.
Further, leveraging activities such as RESAC or derivatives of the National Land Cover Database, also based on Landsat imagery, come at no cost to the Bay Program, or other users.
Bancroft also states that the RESAC maps are “just as inaccurate as the NRI” (National Resource Inventory) estimates of urban change.
This seems unfounded given that, to our knowledge, no such comparisons have been conducted. Doing so would require some care as the NRI relies on a statistical sampling approach designed for regional, not state-scale applications, based on periodic aerial census programs.
In contrast, the RESAC maps have complete “wall to wall” coverage providing fine-resolution information on both the small and larger urban changes that are occurring throughout the Bay watershed. (See accompanying figure.)
While no map is entirely accurate everywhere, and the RESAC maps are least accurate in low-density residential areas where trees can obscure impervious surfaces—even in winter—we believe that these data sets are very suitable for detecting and monitoring change in the built environment and, in fact, provide the best and most practical means of doing so.
We welcome the opportunity to work with the Alliance and others to put this monitoring framework into action, and to support decision making for our common objective—restoring the health and productivity of the Chesapeake Bay.